Regarding emails: if we think that they can be beneficial, but risky, a way to try and hedge against the bad parts is to draft a community agreed upon paragraph that explained the nature of debate and the impact email conversations might have on it. It should also include and explanation that the intent of the conversation is to be used as evidence in a round and that it will not be used unless explicit consent is given.
Include this universal disclaimer along with the questions you ask, and post the emails both asking the questions and the response in a public space and I think you avoid some of the risky things associated with emails.
This would also be verifiable-a simple email to the author asking if the person actually included the disclaimer in their email could quickly confirm or deny any suspicions.
If this were to be accepted the standard would be clear cut and simple: if you include the disclaimer and they acknowledge it, then the email is legit as long as it is posted in a public debate area.
If this system were adopted I think that legitimate email exchanges could occur and we might see more evidence from real experts rather than random sources.
2 other suggestions:
1. A standard for how long in advance of a tournament the evidence must be posted for it to be eligible, so others could have time to correspond with other experts to produce the possible answers necessary.
2. Make it a rule that if you wish to read email evidence, you must also read the entirety of the questions you asked in the previous email during the round, not just have it available if they wish to see it. This would allow for evidence comparisons such as "you are asking leading questions...the author is not advancing this information on their own, just merely agreeing with the question." This would encourage people to ask questions that are less leading and operate more to ask an opinion from a qualified expert when they simply haven't written on an issue.